USING LABORATORY EXPERIMENTS TO TEST THEORIES OF CORPORATE BEHAVIOR

Abstract

Laboratory experiments provide the most rigorous method of testing scientific theories. However, their current use is primarily limited to testing theories of individual behavior. I suggest the conditions under which one can test theories of corporate behavior in laboratory experiments, using human subjects in the role of purposive corporate actors (such as groups or the state). Using the Condorcet Jury Theorem, I demonstrate that, when four conditions are met, laboratory experiments with human subjects represent statistically conservative tests of such theories. I address the issue of `external validity' and argue that it is not a concern for laboratory experiments as a means of testing theories with clearly stated scope conditions. Finally, I point out that, contrary to popular belief, supermajority decision rules, used by juries and legislatures alike presumably for more important decisions, actually lead to inferior collective decisions.